generated: '2026-09-13'
method: probed
source: https://docs.squared.ai/.well-known/mcp.json
name: AI Squared Documentation MCP Server
description: >-
  An anonymous, read-mostly MCP server over the AI Squared documentation corpus, advertised by
  AI Squared at /.well-known/mcp.json on its own documentation host. It exposes documentation
  search, a read-only virtual filesystem over the docs pages and API reference, and a
  docs-feedback write. It is a DOCUMENTATION server, not a control plane: no tool in it calls the
  AI Squared REST API at api.squared.ai, so an agent cannot create a connector, run a sync or read
  sync records through MCP.
status: live
deployment:
  mode: remote
  endpoint: https://docs.squared.ai/mcp
  auth: none
  verified: probed
discovery:
  well_known: https://docs.squared.ai/.well-known/mcp.json
  well_known_status: 200
  advertised_url: https://multiwoven-74.main-kill-isr.mintlify.me/mcp
  note: >-
    The /.well-known/mcp.json manifest names a Mintlify-hosted preview origin
    (multiwoven-74.main-kill-isr.mintlify.me). Both that URL and the provider-branded alias
    https://docs.squared.ai/mcp were POSTed with tools/list; both returned HTTP 200 with an
    identical tool set. The provider-branded endpoint is recorded as the deployment endpoint
    because it is the one on a host AI Squared controls.
x-evidence:
  fetched: '2026-09-13'
  method: POST
  url: https://docs.squared.ai/mcp
  body: '{"jsonrpc":"2.0","id":1,"method":"tools/list"}'
  accept: application/json, text/event-stream
  http_status: 200
  transport: streamable-http (text/event-stream response)
  tools_returned: 3
  raw: ai-squared-mcp-tools.json
authorship: platform
authorship_note: >-
  Generated by the Mintlify documentation platform from AI Squared's own docs, and served under
  AI Squared's host and /.well-known/ manifest. The tool names are namespaced to the provider
  (search_ai_squared, query_docs_filesystem_ai_squared) but the server shape is the vendor's, not
  a hand-built AI Squared server.
tools:
- name: search_ai_squared
  title: Search documentation
  summary: >-
    Search the AI Squared knowledge base for documentation, code examples, API reference entries
    and guides; returns contextual content with titles and direct doc links.
  input_schema:
    query: string (required)
  annotations:
    read_only: true
    destructive: false
    idempotent: true
    open_world: false
- name: query_docs_filesystem_ai_squared
  title: Query the documentation filesystem
  summary: >-
    Run read-only shell-like commands (ls, cat, head, rg, tree) against a sandboxed in-memory
    filesystem containing only the AI Squared documentation pages and API-reference specs.
  input_schema:
    command: string (required)
  annotations:
    read_only: true
    destructive: false
    idempotent: true
    open_world: false
- name: submit_feedback
  title: Submit documentation feedback
  summary: >-
    Report a problem with a documentation page (incorrect, outdated, confusing, incomplete, or a
    broken example) to the docs team. The only write in the server.
  input_schema:
    path: string (required)
    feedback: string (required)
  annotations:
    read_only: false
    destructive: false
    idempotent: false
    open_world: true
platform_mcp_consumption:
  note: >-
    Separately from publishing this server, the AI Squared PRODUCT is an MCP client: its Workflows
    feature ships an Agent component and an MCP Tools component that connect to external MCP
    servers, with 26 pre-integrated remote MCP server tools plus custom MCP server support.
    That is AI Squared consuming MCP, not serving it, and is recorded here so the two are not
    conflated.
  source: https://docs.squared.ai/workflows/mcp-tools